How do you resume a PLC sequence after a light-curtain trip?

When a light curtain trips, the safety function—not the SFC—must stop hazardous motion; the PLC must then preserve or abort the sequence according to the machine's actual state and recovery design. Resume only when measured axis positions, workholding, and process conditions satisfy the next command's preconditions; a remembered active step alone is not proof that continuing is safe. On Allen-Bradley PLC-5 and ControlLogix platforms that support SFCs, restart options include the initial step or last active step, while SFR, SFP, or the PLC-5 SDS instruction can support more deliberate restart selection.

Motion state at the light-curtain trip

A curtain interruption can stop an axis in the middle of a commanded move, before its in-position condition becomes true. A sequencer that waits only for that condition can remain stuck after the safety system has stopped or disabled the motion. The first diagnostic task is to identify where the axis physically stopped and what motion capability remains: a controlled deceleration, a torque-disabled stop, or a shutdown that invalidated the machine reference produce different recovery choices.

The deciding quantity is actual machine state versus the saved step's entry conditions. Read position, velocity or standstill status, drive enable/status, reference validity, and the related process sensors from the motion controller, drive diagnostics, and PLC. Compare them with machine-specific acceptable ranges and interlocks; there is no universal restart tolerance to apply.

Quantity or condition	Recovery limit or decision	Where to read it
Actual axis position	Must be within the saved step's permitted entry window, or a defined recovery move is required.	Motion-controller position feedback and the sequence's saved start/stop positions.
Axis velocity and standstill	Confirm the motion has stopped before selecting a restart path; account for coast or deceleration travel.	Drive or motion-controller status and, where available, its event/diagnostic history.
Drive power, enable, and fault state	Determine whether the drive retained control and whether the safety function permits re-enabling.	Drive status, PLC safety-status interface, and the machine safety circuit.
Position reference or home validity	If the reference was lost or machine position changed independently of feedback, return to a known reference before automatic motion.	Axis reference status, encoder feedback, and machine homing procedure.
Part, clamp, and limit conditions	Every condition required by the next move must still be true, not merely true before the interruption.	Relevant input states, position switches, clamp feedback, and operator inspection.

Record the active sequence step, last motion command, commanded target, measured stopping position, and safety/drive status at the interruption. Those facts distinguish a logic deadlock from a position or drive-state problem. If the sequence advances while the curtain is interrupted, the issue is transition gating; if it remains active but motion cannot resume, inspect the move command and the axis state before changing the step.

How a safety stop creates a sequence-restart problem

An SFC advances when its transition conditions are satisfied. A light-curtain trip is an asynchronous machine event from the sequence's perspective: it may stop motion without completing the normal move and without making the expected in-position bit true. Restarting execution at the same waiting action can therefore reproduce the deadlock. Restarting at the beginning can instead repeat completed operations, move into an occupied space, or invalidate the work already in progress.

Keep the safety function separate from normal sequence control. The curtain and safety circuit determine whether hazardous motion is allowed; ordinary PLC logic determines which production action comes next after the protective condition has been cleared and a deliberate restart is requested. A sequence pause, stored step, feed override, or HMI command cannot replace the safety function or authorize motion while the safety circuit remains tripped.

A zero-percent feedrate or velocity override can be useful as an operational motion hold if the controller supports it. It may preserve the current command and allow continuation after a controlled stop, but it is not automatically a safety-rated stop, and it may leave the drive energized. Evaluate it against the machine's safety design and risk assessment. The safety circuit must still independently achieve the required safe condition.

Recovery path selection for the interrupted move

Choose among resume, recover-and-resume, and abort based on the stopped machine state, not on a preference for preserving the step bit. Resume is appropriate only when the motion controller retains a valid command/reference and the axis and process conditions still match the expected state. Recover-and-resume returns the machine to a defined point or reissues a move from current feedback. Abort clears the interrupted production context and requires the equipment and workpiece to return to known conditions before a new cycle.

Observed condition	Candidate response	Decision check
Drive stopped in a known position; command and reference remain valid	Retain the master sequence and resume the current move or its waiting action.	Confirm the move can be safely continued and no relevant sensor or clamp state changed.
Move stopped before target, but controller state or command was cleared	Reissue an absolute target, or run a defined recovery move before continuing.	Compare measured position with the target and ensure reissue will not duplicate an incremental displacement.
Axis was moved manually, pushed, or crossed a relevant limit during access	Use a machine-specific recovery path, manual repositioning, or abort and home.	Recheck collision clearances, axis references, workholding, and all step preconditions.
Drive power/reference was lost or process state is unknown	Abort automatic continuation; restore known state through the approved recovery procedure.	Verify homing/reference and part condition before starting a new sequence.

Position feedback that survives a stop can preserve axis coordinates, and an absolute encoder can avoid losing position when power is removed. That does not prove that the mechanical process is still correctly referenced: an axis may have been pushed, a coupling may have slipped, or a fixture or part may have moved. Treat retained position as an input to the recovery decision, then validate it against machine and process conditions.

Restart-point logic beyond the active SFC step

The last active step and the last motion command are not necessarily the same recovery point. A motion profile may issue a move first, then trigger several sequence actions as position changes. If the stop occurs after one of those actions but before the move completes, restarting at the last active SFC step may wait for a completion bit that will never arrive. Restarting at the motion-command step can reissue the move, but only if the command semantics and current position make that safe.

Absolute and relative moves need different treatment. Reissuing an absolute target normally points back toward the same coordinate, subject to the controller's command behavior and path constraints. Reissuing a relative move can add the increment again and drive beyond the intended destination. Save or reconstruct the intended start and stop positions, use actual feedback to select a recovery target, and explicitly define whether the interrupted move is to finish, return to its start, or be abandoned. Relative moves become especially sensitive to interruption because the reference point may have changed.

For the Allen-Bradley platforms identified above, SFC restart configuration can select the initial step or last active step. The SFR and SFP ladder instructions provide a way to halt SFC execution and restart at a selected step; the PLC-5 SDS instruction was used to analyze relevant inputs and determine a restart point. These mechanisms select execution state, not a safe physical trajectory. The application still needs logic that determines whether the selected step is valid from the measured axis and process state.

SFC behavior and restart details vary by implementation. Verify the target controller's supported instructions, restart configuration, retentive behavior, and scan/transition behavior in its programming documentation. Do not assume that an SFC pause or interrupt automatically rewinds to the most recent motion command.

Sequence architecture that retains state without duplicating every step

Use a retained master sequence to represent the machine-level operation and a distinct pause/abort decision to govern whether transitions may occur. A common pattern is to gate every normal state transition with a shared continue-enable condition. When the safety circuit trips, the sequence stops advancing while the safety function independently stops motion. On a validated resume request, the master sequence can continue from its saved state or transfer to an explicit recovery branch.

This avoids adding a separate pause step after every ordinary step, which can nearly double the number of states and create mismatched transition logic. A shared transition gate does not mean every motion can be resumed identically: each move still needs its own recovery rule. Some moves can finish at a reduced operational speed; others must return to the move's start or abort. Keep those decisions local to the move or its slave sequence while the master retains cycle context.

Define pause and abort as different outcomes. A pause preserves enough state for a qualified recovery attempt and blocks normal step transitions. An abort clears or invalidates the interrupted cycle and requires return to known conditions before starting again. Some faults warrant pause; others—such as an invalid position reference or a process condition that cannot be reconstructed—must force abort. Specify what happens to ordinary outputs and motion commands in each state rather than relying on a generic paused bit.

Multi-axis recovery with master, slave, or virtual-axis control

For coordinated equipment, retain a master sequence for machine progression and use per-axis slave sequences to handle the physical recovery of each axis. A slave can decide whether to complete the current move slowly, return to its saved start position, or wait for other axes. A master can store each step's start and stop positions and wait until the required axes are synchronized before allowing the coordinated move to run again. The exact order depends on the machine's collision envelope and the process path.

A virtual axis is another option where the motion platform supports it. In the described ControlLogix approach, the virtual axis stops on the safety trip while physical slave axes are cammed to it; during recovery, the virtual trajectory and cam relationship guide the slaves toward the restart position. This can provide a common progression reference, but it adds configuration and commissioning work, including management of cam data. It is not a shortcut around defining safe restart positions.

Complex paths need an explicit escape or recovery trajectory. A CNC tool moving around a workpiece, or several interlocking axes, may not be able to reverse along the original path without crossing the part or another mechanism. Determine recovery ordering, collision clearances, and synchronization for each possible stopped region. A generic instruction to back up to the previous state is insufficient when the safe path depends on where every axis stopped.

Commissioning a controlled resume procedure

Implement and test recovery as a defined operating path, not as an automatic consequence of clearing the curtain. Use this sequence during design and commissioning:

1. Define stop behavior. Document what the safety system does to each drive on curtain interruption, including whether motion decelerates or drive power/enable is removed. Confirm the safety function itself independently of the sequence code.
2. Capture interruption state. Latch the active machine step, active axis move, target, actual positions, drive status, reference validity, relevant sensor states, and whether the interrupted product remains recoverable.
3. Classify the event. Route it to pause, recover-and-resume, or abort based on explicit conditions. If a reference is invalid or a required condition cannot be proven, block automatic continuation.
4. Revalidate before motion. After the curtain is clear and the safety circuit permits reset, require the intended operator restart action and check the entry conditions for the selected recovery step. A restored curtain alone must not trigger motion.
5. Issue the selected recovery move. Reissue only a command whose target and semantics remain valid; otherwise move to a defined safe start or require homing/manual recovery. Use a reduced operational speed only where the controller and machine design permit it.
6. Confirm completion and hand back control. Verify actual position and process feedback, then permit the master sequence to advance. Clear the pause latch only after the axis and sequence agree on the same recovery state.

Make the operator interface identify whether the machine is paused, recovering, aborted, or ready for a new cycle, and show the axis or condition preventing continuation. The operator should not need to guess whether the current move will resume, restart, or be discarded.

Verification tests for each recovery branch

Test the normal interruption point that exposes the original problem: trip the curtain during a move before the in-position condition, then confirm motion stops through the safety design while the production sequence does not falsely complete or advance. After clearing the protective condition and issuing the designated reset/start actions, verify that the chosen recovery branch reissues or reconstructs the motion rather than waiting forever on a stale completion bit.

Repeat the test at points where the move has crossed a position-triggered transition, where one axis has stopped farther along than another, and where an operator may have changed a mechanism or workpiece during access. Check that the sequence rejects a stale step-complete bit when its underlying condition is no longer true. Test both pause and abort paths, including loss of reference or drive fault, and verify that abort requires the defined return to known conditions.

Record actual stopping position and stop time from the drive/controller diagnostics during commissioning, then compare them with the machine's restart windows and clearances. Confirm the safety reset does not itself restart automatic motion. After every test, verify axis reference, clamps, limits, product location, and the master/slave sequence state before returning the machine to production.

Failure modes that make a clean resume unsafe

A saved step can contain assumptions that became false while the sequence was paused. An earlier step may have confirmed a clamp, limit, or product position, but a later interruption can allow that condition to change. Recheck the conditions at the moment of restart instead of treating a past transition as permanent authorization. A newly crossed limit or moved part can make the old path invalid even when the PLC still shows the expected step.

Manual mode creates the same hazard. An operator may jog an axis or reposition a fixture during access and then return to automatic. Check actual coordinates and process state against the saved move before resuming; do not let an auto-mode change alone restore motion outputs. For a machine where recovery depends on homing, coordinate preservation is only useful if the encoder-to-mechanism relationship remained valid.

Feed override and virtual-axis recovery can reduce programming complexity in suitable applications, but neither removes the need for step-specific trajectory analysis. A zero-speed command can leave torque active, and a cammed restart can still take a slave through an unsafe path if the reference or cam relationship is wrong. Validate motion behavior with the actual axis configuration and ensure an independent safety function remains responsible for personnel protection.

Finally, treat vendor SFC settings as implementation details rather than portable language semantics. Check the controller's restart mode and instruction behavior before relying on initial-step, last-active-step, or selected-step restart. A design that is understandable to future maintainers should expose the saved state, recovery choice, and reason an automatic resume is blocked.

FAQ: PLC sequence recovery after a light-curtain trip